--- title: "Exploratory Data Analysis" author: "Pranav Patil" format:   html:     embed-resources: true ---

Importing Libraries

import pandas as pd
import plotly.express as px
import matplotlib.pyplot as plt
import seaborn as sns

Loading Dataset

data = pd.read_csv("data_final.csv")

print(data.head())

   Unnamed: 0       WHO Region ISO3 WHO Country Name    City or Locality  \
0           0  European Region  ALB          Albania             Elbasan   
1           1  European Region  ALB          Albania             Elbasan   
2           2  European Region  AND          Andorra  Escaldes-Engordany   
3           3  European Region  AND          Andorra  Escaldes-Engordany   
4           4  European Region  AND          Andorra  Escaldes-Engordany   

   Measurement Year  NO2 (μg/m3)  NO2 temporal coverage (%)  Population  \
0              2015        23.96                  97.853881   2880703.0   
1              2016        26.26                  96.049636   2876101.0   
2              2012        31.64                 100.000000     71013.0   
3              2014        27.62                 100.000000     71621.0   
4              2015        26.65                  94.554795     71746.0   

   Average NO2  
0        23.96  
1        26.26  
2        31.64  
3        27.62  
4        26.65  

Summary statistics of Dataset

print(data.describe())

         Unnamed: 0  Measurement Year   NO2 (μg/m3)  \
count  16364.000000      16364.000000  16364.000000   
mean    8181.500000       2015.587570     20.118549   
std     4724.024238          2.722397     11.370591   
min        0.000000       2010.000000      0.000000   
25%     4090.750000       2014.000000     11.857500   
50%     8181.500000       2016.000000     18.505000   
75%    12272.250000       2018.000000     26.710000   
max    16363.000000       2021.000000     89.770000   

       NO2 temporal coverage (%)    Population   Average NO2  
count               16364.000000  1.636400e+04  16364.000000  
mean                   96.802395  1.433897e+08     20.118549  
std                     2.648411  3.385217e+08      4.475790  
min                    90.011000  3.702900e+04      1.080000  
25%                    94.931507  1.958872e+07     16.986737  
50%                    97.350000  5.972908e+07     20.023105  
75%                    99.132420  6.715835e+07     23.060341  
max                   100.000000  1.383112e+09     59.820000  

Scatter plot for NO2 levels over the years

fig = px.scatter(data, x="Measurement Year", y="NO2 (μg/m3)", color="WHO Country Name",
                 title="NO2 Levels Over the Years", labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)"})
fig.show()

Bar chart for average NO2 levels by country

avg_no2_by_country = data.groupby("WHO Country Name")["Average NO2"].mean().reset_index()
fig = px.bar(avg_no2_by_country, x="WHO Country Name", y="Average NO2",
              color="WHO Country Name", title="Average NO2 Levels by Country", 
             labels={"Average NO2": "Average NO2 Levels (μg/m3)"})
fig.show()

Bar chart for average NO2 levels by region

avg_no2_by_region = data.groupby("WHO Region")["Average NO2"].mean().reset_index()
fig = px.bar(avg_no2_by_region, x="WHO Region", y="Average NO2",
              color="WHO Region", title="Average NO2 Levels by Region", 
             labels={"Average NO2": "Average NO2 Levels (μg/m3)"})
fig.show()

Bar chart for average NO2 levels by city

avg_no2_by_city = data.groupby("City or Locality")["Average NO2"].mean().reset_index()
fig = px.bar(avg_no2_by_city, x="City or Locality", y="Average NO2",
              color="City or Locality", title="Average NO2 Levels by City", 
             labels={"Average NO2": "Average NO2 Levels (μg/m3)"})
fig.show()

Heatmap for NO2 levels by country and year

fig = px.imshow(data.pivot_table(values="NO2 (μg/m3)", index="WHO Country Name", columns="Measurement Year"),
                labels={"index": "Country", "columns": "Year", "value": "NO2 Levels (μg/m3)"},
                title="NO2 Levels by Country and Year")
fig.show()

Heatmap for NO2 levels by region and year

fig = px.imshow(data.pivot_table(values="NO2 (μg/m3)", index="WHO Region", columns="Measurement Year"),
                labels={"index": "Region", "columns": "Year", "value": "NO2 Levels (μg/m3)"},
                title="NO2 Levels by Region and Year")
fig.show()

Heatmap for NO2 levels by city and year

fig = px.imshow(data.pivot_table(values="NO2 (μg/m3)", index="City or Locality", columns="Measurement Year"),
                labels={"index": "City or Locality", "columns": "Year", "value": "NO2 Levels (μg/m3)"},
                title="NO2 Levels by City and Year")
fig.show()

Heatmap for NO2 levels by city and year in Eastern Mediterranean Region

# Filtering data for Eastern Mediterranean Region
eastern_med_data = data[data['WHO Region'] == 'Eastern Mediterranean Region']

fig = px.imshow(eastern_med_data.pivot_table(values="NO2 (μg/m3)", index="City or Locality", columns="Measurement Year"),
                labels={"index": "City", "columns": "Year", "value": "NO2 Levels (μg/m3)"},
                title="NO2 Levels by City and Year in Eastern Mediterranean Region")
fig.show()

Heatmap for NO2 levels by city and year in European Region

# Filtering data for European Region
eu_data = data[data['WHO Region'] == 'European Region']

fig = px.imshow(eu_data.pivot_table(values="NO2 (μg/m3)", index="City or Locality", columns="Measurement Year"),
                labels={"index": "City", "columns": "Year", "value": "NO2 Levels (μg/m3)"},
                title="NO2 Levels by City and Year in European Region")
fig.show()

Heatmap for NO2 levels by city and year in Region of the Americas

# Filtering data for Region of the Americas
reg_amer_data = data[data['WHO Region'] == 'Region of the Americas']

fig = px.imshow(reg_amer_data.pivot_table(values="NO2 (μg/m3)", index="City or Locality", columns="Measurement Year"),
                labels={"index": "City", "columns": "Year", "value": "NO2 Levels (μg/m3)"},
                title="NO2 Levels by City and Year in Region of the Americas")
fig.show()

Heatmap for NO2 levels by city and year in South East Asia Region

# Filtering data for South East Asia Region
se_asia_data = data[data['WHO Region'] == 'South East Asia Region']

fig = px.imshow(se_asia_data.pivot_table(values="NO2 (μg/m3)", index="City or Locality", columns="Measurement Year"),
                labels={"index": "City", "columns": "Year", "value": "NO2 Levels (μg/m3)"},
                title="NO2 Levels by City and Year in South East Asia Region")
fig.show()

Heatmap for NO2 levels by city and year in Western Pacific Region

# Filtering data for Western Pacific Region
western_pac_data = data[data['WHO Region'] == 'Western Pacific Region']

fig = px.imshow(western_pac_data.pivot_table(values="NO2 (μg/m3)", index="City or Locality", columns="Measurement Year"),
                labels={"index": "City", "columns": "Year", "value": "NO2 Levels (μg/m3)"},
                title="NO2 Levels by City and Year in Western Pacific Region")
fig.show()

Bubble plot for NO2 levels by city and year with bubble size based on population for Eastern Mediterranean Region

fig = px.scatter(eastern_med_data, x="Measurement Year", y="City or Locality", size="Population",
                 color="NO2 (μg/m3)", hover_name="City or Locality",
                 title="NO2 Levels by City and Year in Eastern Mediterranean Region",
                 labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Population": "Population"},
                 size_max=30)  
fig.show()

Bubble plot for NO2 levels by city and year with bubble size based on population for European Region

fig = px.scatter(eu_data, x="Measurement Year", y="City or Locality", size="Population",
                 color="NO2 (μg/m3)", hover_name="City or Locality",
                 title="NO2 Levels by City and Year in European Region",
                 labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Population": "Population"},
                 size_max=30)
fig.show()

Bubble plot for NO2 levels by city and year with bubble size based on population for Region of the Americas

fig = px.scatter(reg_amer_data, x="Measurement Year", y="City or Locality", size="Population",
                 color="NO2 (μg/m3)", hover_name="City or Locality",
                 title="NO2 Levels by City and Year in Region of the Americas",
                 labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Population": "Population"},
                 size_max=30)  
fig.show()

Bubble plot for NO2 levels by city and year with bubble size based on population for South East Asia Region

fig = px.scatter(se_asia_data, x="Measurement Year", y="City or Locality", size="Population",
                 color="NO2 (μg/m3)", hover_name="City or Locality",
                 title="NO2 Levels by City and Year in  South East Asia Region",
                 labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Population": "Population"},
                 size_max=30)  
fig.show()

Bubble plot for NO2 levels by city and year with bubble size based on population for Western Pacific Region

fig = px.scatter(western_pac_data, x="Measurement Year", y="City or Locality", size="Population",
                 color="NO2 (μg/m3)", hover_name="City or Locality",
                 title="NO2 Levels by City and Year in Western Pacific Region",
                 labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Population": "Population"},
                 size_max=30)  
fig.show()

Choropleth map for NO2 levels by country in each WHO region

fig = px.choropleth(data, 
                    locations="ISO3",
                    color="Average NO2",
                    hover_name="WHO Country Name",
                    animation_frame="Measurement Year",
                    color_continuous_scale=px.colors.sequential.Plasma,
                    projection="natural earth",
                    title="NO2 Levels by Country and Year in Each WHO Region",
                    labels={"Average NO2": "Average NO2 Levels (μg/m3)"})
fig.show()

Violin plot for NO2 levels by Region

fig = px.violin(data, 
                x="WHO Region", 
                y="NO2 (μg/m3)",
                box=True,  # Include box plot inside the violin
                points="all",  # Display individual data points
                color="WHO Region",
                hover_name="WHO Country Name",
                title="Violin Plot of NO2 Levels by Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)"}
               )

fig.show()

fig = px.violin(eastern_med_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                box=True,  # Include box plot inside the violin
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Violin Plot of NO2 Levels by Year and City in Eastern Mediterranean Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

fig = px.violin(eu_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                box=True,  # Include box plot inside the violin
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Violin Plot of NO2 Levels by Year and City in European Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

fig = px.violin(reg_amer_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                box=True,  # Include box plot inside the violin
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Violin Plot of NO2 Levels by Year and City in Region of the Americas",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

fig = px.violin(se_asia_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                box=True,  # Include box plot inside the violin
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Violin Plot of NO2 Levels by Year and City in South East Asia Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

fig = px.violin(western_pac_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                box=True,  # Include box plot inside the violin
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Violin Plot of NO2 Levels by Year and City in Western Pacific Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

Box plot for NO2 levels by region

fig = px.box(data, 
             x="WHO Region", 
             y="NO2 (μg/m3)",
             points="all",  # Display individual data points
             hover_name="WHO Country Name",
             color="WHO Region",
             title="Box Plot of NO2 Levels by Region",
             labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)"}
            )

# Show the plot
fig.show()

fig = px.box(eastern_med_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Box Plot of NO2 Levels by Year and City in Eastern Mediterranean Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

fig = px.box(eu_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Box Plot of NO2 Levels by Year and City in European Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

fig = px.box(reg_amer_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Box Plot of NO2 Levels by Year and City in Region of the Americas",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

fig = px.box(se_asia_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Box Plot of NO2 Levels by Year and City in South East Asia Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

fig = px.box(western_pac_data,
                x="Measurement Year",
                y="NO2 (μg/m3)",
                #color="City or Locality",
                points="all",  # Display individual data points
                hover_name="City or Locality",
                title="Box Plot of NO2 Levels by Year and City in Western Pacific Region",
                labels={"NO2 (μg/m3)": "NO2 Levels (μg/m3)", "Measurement Year": "Year"}
                )

# Show the plot
fig.show()

Pie chart to show the total NO2 levels for each region

# Aggregate data to get total NO2 levels for each region
region_totals = data.groupby("WHO Region")["NO2 (μg/m3)"].sum().reset_index()

fig = px.pie(region_totals, 
             names='WHO Region',
             values='NO2 (μg/m3)',
             title='Total NO2 Levels by Region',
             labels={'NO2 (μg/m3)': 'Total NO2 Levels (μg/m3)', 'WHO Region': 'Region'})

# Show the plot
fig.show()

Group by data EDA

count_df = data.groupby(['Measurement Year', 'WHO Country Name']).size().reset_index(name='Frequency')

# Display the resulting data frame
print(count_df)

     Measurement Year WHO Country Name  Frequency
0                2010          Austria        103
1                2010          Belgium         41
2                2010         Bulgaria         16
3                2010          Croatia          4
4                2010           Cyprus          2
..                ...              ...        ...
339              2019   United Kingdom         98
340              2020           Kuwait          5
341              2020            Qatar          1
342              2020        Singapore          1
343              2021            Qatar          1

[344 rows x 3 columns]

Bar chart to show the total frequency for each country across all years

fig = px.bar(count_df, x='WHO Country Name', y='Frequency', color='Measurement Year',
             title='Frequency of Each Combination of Year and Country',
             labels={'Frequency': 'Count', 'WHO Country Name': 'Country', 'Measurement Year': 'Year'})

# Show the plot
fig.show()

Line plot for the frequency over the years for different countries

fig = px.line(count_df, x='Measurement Year', y='Frequency', color='WHO Country Name',
              title='Frequency of Measurements Over the Years by Country',
              labels={'Frequency': 'Count', 'WHO Country Name': 'Country', 'Measurement Year': 'Year'})

fig.show()

Box plot to show the distribution of frequency across different years

fig = px.box(count_df, x='Measurement Year', y='Frequency',
             title='Distribution of Frequency Across Different Years',
             labels={'Frequency': 'Count', 'Measurement Year': 'Year'})

fig.show()

Scatter plot to show the relationship between measurement year and frequency

fig = px.scatter(count_df, x='Measurement Year', y='Frequency', color='WHO Country Name',
                 title='Scatter Plot of Frequency Over the Years by Country',
                 labels={'Frequency': 'Count', 'WHO Country Name': 'Country', 'Measurement Year': 'Year'})

fig.show()

Heatmap to visualize the distribution of measurements by both year and country

fig = px.imshow(count_df.pivot_table(values='Frequency', index='WHO Country Name', columns='Measurement Year'),
                labels={'index': 'Country', 'columns': 'Year', 'value': 'Frequency'},
                title='Heatmap of Distribution of Measurements by Year and Country')

fig.show()

Choropleth Map of Frequency of Measurements by Country for Each Year

fig = px.choropleth(count_df, 
                    locations='WHO Country Name', 
                    locationmode='country names',
                    color='Frequency',
                    animation_frame='Measurement Year',
                    title='Choropleth Map of Frequency of Measurements by Country for Each Year',
                    color_continuous_scale=px.colors.sequential.Plasma,
                    labels={'Frequency': 'Count', 'Measurement Year': 'Year'})

# Show the plot
fig.show()

Violin plot to visualize the distribution of frequency measurements by year

fig = px.violin(count_df, 
                x='Measurement Year', 
                y='Frequency',
                box=True,  # Include box plot inside the violin
                points="all",  # Display individual data points
                title='Violin Plot of Frequency Measurements by Year',
                labels={'Frequency': 'Count', 'Measurement Year': 'Year'})

# Show the plot
fig.show()

Pie chart to show the distribution of frequencies across countries

fig = px.pie(count_df, 
             names='WHO Country Name',
             values='Frequency',
             title='Distribution of Frequencies Across Countries',
             labels={'Frequency': 'Count', 'WHO Country Name': 'Country'})

fig.show()

region_df = data.groupby(['Measurement Year', 'WHO Country Name', 'WHO Region']).size().reset_index(name='Frequency')

region_df.head()

	Measurement Year	WHO Country Name	WHO Region	Frequency
0	2010	Austria	European Region	103
1	2010	Belgium	European Region	41
2	2010	Bulgaria	European Region	16
3	2010	Croatia	European Region	4
4	2010	Cyprus	European Region	2

Violin plot to visualize the distribution of frequency

fig = px.violin(region_df,
                x='WHO Region',
                y='Frequency',
                box=True,  # Include box plot inside the violin
                points="all",  # Display individual data points
                title='Violin Plot of Frequency Distribution by Region',
                labels={'Frequency': 'Count', 'WHO Region': 'Region'})

# Show the plot
fig.show()

Box plot to visualize the distribution of frequency

fig = px.box(region_df,
             x='WHO Region',
             y='Frequency',
             points="all",  # Display individual data points
             title='Box Plot of Frequency Distribution by Region',
             labels={'Frequency': 'Count', 'WHO Region': 'Region'})

# Show the plot
fig.show()

Bar chart to show the total frequency for each region

total_freq_by_region = region_df.groupby('WHO Region')['Frequency'].sum().reset_index()
fig = px.bar(total_freq_by_region,
             x='WHO Region',
             y='Frequency',
             color='WHO Region',
             title='Total Frequency by Region',
             labels={'Frequency': 'Total Count', 'WHO Region': 'Region'})

# Show the plot
fig.show()

Bubble chart to visualize frequency distribution by region

fig = px.scatter(region_df,
                 x='WHO Region',
                 y='Frequency',
                 size='Frequency',  # Size of the bubble represents the count
                 color='WHO Region',
                 hover_name='WHO Country Name',
                 title='Bubble Chart of Frequency Distribution by Region',
                 labels={'Frequency': 'Count', 'WHO Region': 'Region'})

fig.show()

Pie chart to show the distribution of frequencies across regions

# Aggregate data to get total frequencies for each region
total_freq_by_region = region_df.groupby('WHO Region')['Frequency'].sum().reset_index()

fig = px.pie(total_freq_by_region, 
             names='WHO Region',
             values='Frequency',
             title='Pie Chart of Frequency Distribution by Region',
             labels={'Frequency': 'Total Count', 'WHO Region': 'Region'})

fig.show()

Heatmap for the frequency distribution

fig = px.imshow(region_df.pivot_table(values='Frequency', index='WHO Region', columns='Measurement Year'),
                labels={'index': 'Region', 'columns': 'Year', 'value': 'Frequency'},
                title='Heatmap of Frequency Distribution by Region and Year')

# Show the plot
fig.show()

fig = px.scatter_matrix(data, dimensions=["Measurement Year", "NO2 (μg/m3)", "NO2 temporal coverage (%)", "Population", "Average NO2"])

# Show the plot
fig.show()

/Users/pranavpatil/anaconda3/lib/python3.11/site-packages/plotly/express/_core.py:279: FutureWarning:

iteritems is deprecated and will be removed in a future version. Use .items instead.

fig = px.scatter_matrix(
    data,
    dimensions=["NO2 (μg/m3)", "NO2 temporal coverage (%)", "Population", "Average NO2"],
    color="WHO Region",
    symbol="WHO Country Name",
    labels={"NO2 (μg/m3)": "NO2 Concentration (μg/m3)"},
    title="Pair Plot of NO2 Data",
    height=800,
    width=800
)

fig.update_traces(diagonal_visible=False)

fig.show()

/Users/pranavpatil/anaconda3/lib/python3.11/site-packages/plotly/express/_core.py:279: FutureWarning:

iteritems is deprecated and will be removed in a future version. Use .items instead.